This document presents a control theory view of human performance and its implications for managing performance. It argues that current views of human performance are inadequate and misleading. It proposes that people are "living control systems" that act to control their perceptions of aspects in the world. Performance is influenced by goals, perceptions, discrepancies between goals and perceptions, actions, and other conditions. For managing complex performance where results are indirect, managers should focus on goals, perceptions, distinguishing between behavior and performance, and enabling effective interventions, rather than trying to control behavior. The key is helping performers frame interventions that work through the situation's structure to achieve desired results.

1. A Control Theory View
of Human Performance
Fred Nickols
© Fred Nickols 2006
All rights reserved
www.nickols.us
nickols@att.net

2. A Control Theory View of Human Performance
Purpose
In the first part of this paper I will present a view of human performance that is rooted in control
theory as articulated by William T. Powers in his seminal work, Behavior: The Control of Per-
ception and as elaborated upon in later publications1. In the second part of this paper I will use
this control theory view to provide insights into the nature of human performance in the work-
place and to draw attention to the challenges this view presents to those who wish to engineer or
manage the performance of others.
Why?
Why write a paper about yet another view of human performance? In particular, why write one
from a control theory perspective? The simple answer is that current views are inadequate and,
more important, incorrect, which means they are misleading and counter-productive; in a word,
they are dysfunctional. Generally speaking, the current view of human performance is that it can
be engineered, that it can be managed, and that it can be engineered and managed by people other
than the performer(s) in question (e.g., by industrial engineers on the one hand and by managers
and executives on the other). The current view is clearly valid in situations in which performance
consists of producing well-defined, tangible results using predetermined tools, methods and tech-
niques. It is less valid in situations in which the performer must configure or tailor his or her re-
sponse to the situation at hand, especially when the situation itself is in a state of flux. Something
else is required for these situations. In my opinion, that “something else” is a control theory view
of human performance. Such a view has implications for performance and performance man-
agement. Because it is clearly the case that the chief means of realizing human performance is
human behavior, let us begin there, with a control theory view of human behavior.
Living Control Systems
People are “living control systems” to use William T. Powers’ term; that is, they act to control
certain aspects of the world about them. All that any of us know of the world about us we know
via our perceptions. What we in fact act to control, then, are our perceptions of certain aspects of
the world about us. Hence, Powers’ view that behavior is the control of perception.
Control is always against some standard, a reference, a goal. In Figure 1 below, the concept of a
goal, a standard, a preferred or required condition, is represented by G. In the technical parlance
of control theory, this is known as the reference condition or reference signal, but we will use
simpler, everyday language here.
Our goals or reference conditions are compared with our perceptions of those things to which our
goals or reference conditions apply. In the workplace, a goal or reference condition (i.e., some
aspect of our world that we are attempting to control) might relate to cost and our goal or refer-
ence condition might be to reduce cost by some percentage. We can compare reports and calcula-
tions of actual costs with our goal or reference condition. In an engineered control system, a
comparator performs this function; in a living control system, the comparator is the human being
at its center, the performer.
1
See the Recommended Reading section at the end of this paper.
© Fred Nickols 2006 2

3. A Control Theory View of Human Performance
In Figure 1, the small black circle on the left, where G and P come together, represents the com-
parator or the function of comparing perceived conditions (P) with goal (G) or reference condi-
tions. Any difference between the goal condition and the perceived condition constitutes a gap or
discrepancy. In Figure 1, this gap or discrepancy is represented by the letter d. In control the-
ory, any difference between the reference condition and the perceived condition constitutes an
error. In ordinary conversation, “error” has certain negative connotations and might be mislead-
ing in situations in which the discrepancy constitutes better than hoped for results instead of
worse than hoped for results. For example, if the goal were to reduce costs by 15% and current
conditions indicate they have been reduced by 20%, it is difficult to view that as an “error” except
in the strict sense of control the-
ory. So, for our purposes, we will
refer to differences between goal
conditions and perceived condi-
tions as “discrepancies,” not as
“errors.” A lower-case d on the
line leading into A is used to
indicate a discrepancy.
Control theory tells us that errors,
or discrepancies between goal or
reference conditions and per-
ceived conditions, produce action
(A). Effective action, in turn,
affects that which we are at-
tempting to control. In reducing
costs, substituting less expensive Figure 1:
materials, reducing or eliminating
waste and making processes more
efficient are all actions that will A Control Theory View of Human Behavior and Performance
have the desired effect on costs. (The GAP-ACT Model)
But we, and those targeted
variables (T) we seek to control, whether they are cost or revenue or productivity or whatever, do
not exist in isolation. There are other actors and factors at work. There are other influences oper-
ating on the same variables we seek to control. Suppliers raise their prices; the cost of material
goes up. Union negotiations result in pay raises and the cost of production goes up (and goes
right down again as the result of layoffs, reengineering, and certain process improvements). In
control theory, these other influences are called “disturbances,” meaning they disturb the value of
the variables we are trying to control. For our purposes we will simply refer to them as condi-
tions, represented in Figure 1 by the letter C.
In general, then, the model in Figure 1 suggests that behavior, in the context of workplace per-
formance, operates as follows:
• As performers, we seek to control the value of some variable in the workplace (T). That
might be cost, productivity level, sales volume or any of innumerable other variables. At
any point in time, the variable we seek to control has some current value and we refer to
this as the “current condition.”
© Fred Nickols 2006 3

4. A Control Theory View of Human Performance
• What we know of this targeted or “controlled variable (CV)” to use the proper term from
control theory, we know through our perceptions and it is this “perceived condition” (P)
that we compare with our goal (G) or reference condition.
• Discrepancies between our goal or reference conditions for a variable and the condition
we perceive lead to action (A).
• Our actions are helped or hindered by other conditions (C), to a greater or lesser degree,
and the effects of our actions on the controlled variable are the sum of our actions and
these other conditions. Our actions are also part of us and part of the situation in which
and on which we are acting. This is reflected in the overlapping of the circles in Figure 1.
We human beings are by nature goal seeking organisms, “living control systems,” to use Powers’
term. We act to control certain aspects of the world about us and, at the same time, other actors
and factor are often acting to control those same aspects of our world, often in ways and direc-
tions that are at odds with our intentions. Enter management.
Management Implications of a Control Theory View of Human Performance
The structure of the diagram in Figure 1 suggests some obvious leverage points and likely strate-
gies regarding performance management. A brief discussion of these is next.
Distinguish between Behavior and Performance
Perhaps the most important thing to do as part of any effort to manage performance is to draw a
distinction between behavior and performance. Behavior refers to the activity of the performer;
performance refers to the effects or outcomes of that activity. To illustrate, consider the tracking
task that is frequently cited in discussions of perceptual control theory. In this task a person is
given a joystick that controls the movement of a cursor on a screen. Also on the screen is a target
symbol, one that moves about. The task requires the subject to keep the cursor aligned with the
target. The subject’s observable behavior consists mainly of moving the joystick; the perform-
ance is the alignment between the cursor and the target. Were you to ask from a behavioral per-
spective, “What is this person doing?” you might well answer, “Moving the joystick.” But from a
performance perspective the answer would be, “Keeping the cursor and the target aligned.” It is
well to keep in mind that the aim of performance management is not the control of human behav-
ior but the reliable attainment of results and this depends on making a distinction between behav-
ior and performance.
Focus on the Goal State
A moment’s reflection will make clear that, except by chance, it would be impossible for an em-
ployee – a performer – to perform as expected unless the employee holds in mind the proper goal.
More important, “holding” this goal entails more than merely comprehending it. For a goal to be
operative at least five conditions must be met:
1. Clarity. The performer must have a clear understanding of the goal; a murky, muddy,
fuzzy grasp of what is expected or required won’t lead to attainment of the required re-
sults. Obviously, two-way communication is called for.
2. Consistency. The performer’s understanding of the goal must be consistent with the ex-
pectations of management regarding the goal; otherwise, some other result is likely to be
achieved. Again, two-way communication is essential.
© Fred Nickols 2006 4

5. A Control Theory View of Human Performance
3. Consequences. The consequences of success should be such that the performer has rea-
son to pursue the goal and the consequences of failure should provide the same incentive.
Two types of consequences are of interest here: natural, and contrived. The natural con-
sequences are those inherent in achieving the goal, be they viewed as positive or negative
by the performer. Contrived consequences are those established by management as a
way of inducing the performer to pursue this or that goal. The wise manager understands
that the real task of managing consequences isn’t one of simply imposing contrived con-
sequences but, instead, is a matter of mitigating the natural consequences. To be sure,
contrived consequences can play a role but a manager should never lose sight of the natu-
ral consequences of actions and should always view contrived consequences in light of
their interplay with those natural consequences. Failure to integrate the two can result in
situations in which the natural and contrived consequences of action do little except nul-
lify one another. Incentive structures, then, must always consider not just the balance of
positive and negative consequences but the balance of natural and contrived conse-
quences as well.
4. Conflict. Closely related to the issue of consequences is the issue of goal conflict. Rare
is the situation in which a performer pursues a single goal. Most people in organizations
have several goals and objectives “on their plate” so to speak. People are often asked to
pursue goals that they see as conflicting; reduce costs and improve productivity; effect a
staff cut and improve staff morale; improve quality and assign your best people to special
projects. Some of this conflict is imagined or attributable to a limited view of the means
available for obtaining the ends in question but some of it is very real. It is part of the
manager’s responsibility to minimize conflict between and among the goals a given per-
former is expected to achieve.
5. Commitment. In the last analysis, the issues above coalesce into a single issue: the per-
former’s commitment to the goal in question. Without this commitment on the part of the
performer, achievement of the goal is doubtful. The issues above, then, should be viewed
as checkpoints in working toward that commitment.
Ensure Accurate, Timely Perceptions of the Current Condition
The diagram shown in Figure 1 is essentially a closed-loop, negative-feedback model. As actions
take effect and are perceived as closing any gap between the goal state and perceived condition,
the need for action is reduced. Once the gap has been closed or the result achieved, efforts shift
from acquisition to maintenance, from closing the gap to keeping it closed. It follows that the
performer’s perceptions of the current state are a critical component of any performance man-
agement system and that it is essential for the performer to have timely, accurate information
about the current state of the condition he or she is attempting to control. This means measure-
ment and communication. Few things are more frustrating than to saddle a performer with the
responsibility for reducing costs in an operation and then withholding information about costs.
Absurd as that might sound, it happens. Worse, to reduce costs almost always entails examining
what is referred to as the cost structure of the operation; that is, how costs are distributed and in-
curred. This information is also often withheld – on the basis of being “too sensitive” for sharing.
Most important, that kind of information is essential to formulating actions that will have the de-
sired effects. With that in mind, let us move on to what many will view as the central issue: ac-
tions.
© Fred Nickols 2006 5

6. A Control Theory View of Human Performance
Distinguish between Prefigured and Configured Actions
At the extremes, there are two kinds of performances expected of people in organizations: (1)
those in which the results of interest are direct and immediate (e.g., as in the case of someone as-
sembling electronic components into a finished product such as a personal computer) and (2)
those in which the results of interest are indirect and delayed (e.g., as in the case of a manager
attempting to turn around a failing division or department). In the first case, because the desired
results are essentially the direct and immediate effects of observable behavior, the management of
performance is mainly a matter of paying attention to the basics of what might be termed “Per-
formance Management 101”:
• Clearly communicate expectations of performance to the performer;
• Provide appropriate incentives (and be sure to remove any inappropriate disincentives);
• Select, train and develop the performer so as to be able to perform as expected
• Provide timely feedback regarding actual performance in relation to expected perform-
ance;
• Remove obstacles and barriers (i.e., those “other influences”) that impede or prevent per-
forming as expected.
These basics of performance management are reasonably well known and, at first glance, com-
patible with a control theory view of human performance. As a matter of fact they work reasona-
bly well with the first kind of performance. This is because the performance in question is both
observable and repetitive. The necessary actions can be studied over time and reduced to a pre-
figured routine. The fabled “one best way” can be determined. However, attempting to apply
this approach to the second kind of performance is rife with risk and fraught with failure. Instead
of thinking primarily in terms of the performer’s observable behavior or immediate actions, think
instead in terms of those actions as initiating interventions. Here’s why
When the results sought (i.e., the conditions intended to be controlled) are indirect and delayed,
the observable behaviors or overt actions of the performer in question do not by themselves ac-
count for those results. Instead, the performer’s actions are best viewed as interventions in the
structure of the situation at hand.2 Moreover, these actions are taken in various places at various
times and their effects are felt in other places at other times. In other words, the direct and imme-
diate effects of actions taken by the performer must “ripple through” the structure of the situation,
making themselves felt elsewhere.
From the preceding, three requirements for successful intervention can be identified:
1. Point(s) of Evaluation. The “elsewhere” mentioned above must be identified. Just
where is it (and when is it) that the desired results are to be gauged? Where, when and
how will the goal state be assessed?
2. Structure. As used here, this term refers to the elements, connections and relationships
between and among those elements that define the structure of the situation in which per-
formance is to be manifested. In which ones must change must be effected and felt?
What are the paths by which changes will “ripple through” this structure, making them-
selves felt at the appropriate points?
2
By “intervention” I mean patterned, purposeful actions, actions taken with an outcome in mind.
© Fred Nickols 2006 6

7. A Control Theory View of Human Performance
3. Point(s) of Intervention. Which of these elements and relationships are accessible and
subject to manipulation by the means available to the performer? Which can be changed
in ways that produce the desired effects (without any off-setting side effects)? What is
the “solution path” (i.e., the path from the points of intervention through the structure of
the situation to the points of evaluation)?
When it comes to the second kind of performance, the kind that requires the performer to config-
ure his or her response, it is far more important from a managerial perspective to focus on helping
the performer frame and carry out effective interventions than on trying to control the behavior of
the performer. (Presumably, although this is by no means a safe assumption, managers have a
broader, deeper understanding of the structures in which interventions must be effected in order
to achieve the desired organizational results.)
It is worth mentioning at this point that Powers’ “living control systems” are also profitably
viewed as continuous problem solvers. They close gaps and they keep them closed. Said a little
differently, people regularly engineer solutions to a wide array of constantly changing problems
on an ongoing basis. In other words, they configure their actions instead of simply carrying out
prefigured routines. When attempting to manage performance, you must make a clear distinction
between performances that hinge on prefigured routines and those that depend on configured rou-
tines because your approach to performance management will be very different.
Manage Those “Other Influences”
Just as the best-laid plans can go astray, the best-framed interventions can fail to produce the de-
sired results. There are other influences at work. In the case of the comparatively simple, pre-
figured assembly line work routine mentioned earlier, performance can be disrupted in many
ways; by a missing tool or part, by a malfunctioning piece of test equipment, by a disruption up
the line and by uncontrolled, unmanaged interruptions and other “disturbances.” Again, owing to
their visibility and immediacy, these kinds of disturbances or disruptions are far more manageable
than the kinds of “other influences” that come into play with configured actions or interventions.
But consider the head of a business unit, say a division, charged with improving the profitability
of the division so as to enhance the return on equity (ROE) of the division. As just one example
of the kinds of structures with which this person must concern himself (or herself) and in which
he or she must intervene, take a look at Figure 2. It illustrates the mathematical-financial struc-
ture of the division’s accounting system, particularly, its calculation of return on equity.3
The structure or system shown in Figure 2 ties in turn to numerous operating systems and struc-
tures and those tie in their turn to social and behavioral structures and systems involving people.
Framing and executing successful interventions in this situation requires being able to map these
three kinds of structures, including their linkages, determining appropriate points of evaluation,
identifying suitable points of intervention and tracing the paths from the latter to the former.
Now the kinds of disturbances or other influences are themselves much less direct and much less
immediate. Political opposition to the proposed changes (overt and covert); failure to adequately
map the structures and linkages so that actions taken have severe unintended consequences and
generate backlashes and new problems; half-hearted cooperation versus energetic collaboration;
corporate “snipers” waiting in the wings to do in the attempted changes (not to mention those at-
tempting to effect them).
3
It should be pointed out that Figure 2 illustrates just one way of calculating ROI. Numerous variations
exist and one should always be sure to use a model of the calculation actually in use.
© Fred Nickols 2006 7

8. A Control Theory View of Human Performance
Gross Sales
Net
minus
Sales
Discounts, Allowances & Returns
Gross
minus
Profit
Materials Cost
Cost of
Manufacturing Labor Cost
Sales
Production Overhead
Net
Promotion minus
Selling Profit
Advertising
Expense
Compensation
plus
Rent & Utilities
Admin
Office Supplies Operating
Expense
Administration Expense
plus
Payroll Taxes Return
Other
Depreciation on
Expense
Insurance Equity
Cash
Inventory Current
Receivables Assets
Prepaid Expenses
Total
plus
Assets
Plant Fixed
Equipment Assets
minus Equity
Notes Payable
Current
Interest Payable
Liabilities
Accounts Payable
Total
plus
Liabilities
Mortgage Fixed
Notes & Bonds Liabilities
Figure 2 - The Structure of Return on Equity (ROE)
To be sure, these other actors and factors can be managed to some extent. Alliances can be made,
sources of opposition can be defused or neutralized, known side effects can be compensated for
and contingency measures established to deal with unforeseen side effects. But it is a far simpler
task to ensure a supportive, relatively disturbance-free work environment for that electronic as-
sembly line worker than it is for the head of a business unit or any other person whose work re-
quires them to configure and carry out interventions into ongoing, dynamic structures.
Support Performance Instead of Controlling the Performer’s Behavior
Thus far, attention has centered mainly on the situation, on actions, other influences, and the ef-
fects of both on the current or controlled condition. Attention paid the performer has centered on
establishing and developing commitment to the goal state. Let us now consider the performer.
Just as conventional performance management copes fairly well with prefigured performances,
conventional practices with respect to the performer are well known, widely understood and rea-
sonably productive. We know that the performer must have an appropriate repertoire or skill set;
we know that getting the right person in the right slot is an important aspect of performance man-
agement; and we know that the selection, training and development of people can contribute
greatly to improved performance. Where we go astray is in not recognizing that the more the
results of interest are indirect and delayed with respect to the performer’s actions, the more we
must be concerned with the situation, not the performer. Yet, we wrongly persist in applying
© Fred Nickols 2006 8

9. A Control Theory View of Human Performance
tools, methods and techniques that work well with prefigured performances to configured per-
formances where they are not only sure not to work but to be counter-productive as well.
I can, for example, prescribe a behavioral routine for many workers whose work performance
yields direct, immediate results. I can also be reasonably confident in my ability to ensure com-
pliance with such a prescription. I can get what I want at a tolerable cost. Chances are, I cannot
successfully develop let alone impose a behavioral prescription for someone whose work per-
formance produces indirect, delayed results. Further, attempting to do so is sure to engender seri-
ous conflict between me and the performer in question. I must take a different approach. I must
enlist commitment to the organization’s goals; I must assist the performer in framing and in car-
rying out effective interventions; I must ensure the availability of timely, accurate information
about the results of actions taken; and I must manage those “other influences.” In short, I must
support the performer in pursuit of the desired performance instead of attempting to control the
performer’s behavior. You see, whenever I attempt to control another person’s behavior I am
interfering with the means by which that person attains all the ends he or she seeks, not just those
that define the performance in question. Not only does conflict ensue, I am guilty of focusing on
means instead of ends and that is often a fatal mistake in management, especially the management
of performance.
Conclusion
To manage performance requires that a distinction be made between behavior and performance.
The former refers to the activity of the performer; the latter refers to the outcomes or effects of
that activity. To manage performance also requires that managers focus on supporting the per-
former’s pursuit of the desired results instead of simply trying to control the performer’s behav-
ior. People, after all, are living control systems and their behavior is the chief means by which
they attain all the goals they seek, not just those that define the performance for which they are
responsible. To support performance, several basic strategies should be used:
• Distinguish between behavior and performance
• Focus on defining and developing commitment to the goal state
• Ensure accurate, timely perceptions of current conditions
• Distinguish between prefigured and configured actions
• Think in terms of interventions
• Manage those “other influences”
• Support performance instead of controlling behavior
References
1. Powers, W. T. (1973). Behavior: The Control of Perception. Aldine De Gruyter: New
York, NY.
Recommended Reading
1. Powers, W. T. (1989). Living Control Systems: Selected Papers of William T. Powers.
The Control Systems Group: Gravel Switch, KY.
2. Marken, Richard S. (1995). Mind Readings: Experimental Studies of Purpose. New
View Publications: Chapel Hill, NC.
3. Powers, W. T. (1998). Making Sense of Behavior: The Meaning of Control. Benchmark
Publications: New Caanan, CT.
© Fred Nickols 2006 9